Huemer is wrong about skepticism (Part 1)

Part 1

If you enjoy this article, please hit the “Like” button (❤️) at the beginning or end of the article. This helps others witness the glorious destruction of moral realism.

---

In this series, I will be critiquing the claims Huemer makes about skepticism in this post. It is probably a good idea to read that post before continuing.

1.0 Introduction

In a recent blog post, Mike Huemer purports to debunk moral skeptics, i.e., people who doubt there are stance-independent moral facts or that we can have knowledge of them.

I don’t find any of what Huemer says convincing, and I don’t think you should, either. Huemer’s arguments rely on helping himself to assumptions I don’t accept and don’t think others have good reasons to accept. Huemer’s objections also amount to a kind of self-defeat argument that relies on vague, speculative armchair psychology about the potential motives and inclinations of moral skeptics. Psychologizing one’s opposition is a fraught enterprise (which isn’t to say psychological hypotheses can’t be correct and relevant to evaluating a philosophical view; I certainly think they can be since I use them myself). Some of what Huemer says seems plausible to me, and some doesn’t, but ultimately, Huemer has in mind a rather tepid and generic skeptic who, even if they are vulnerable to his objections, aren’t representative of other, stronger objections to his views and the views of other rationalistically inclined analytic philosophers.

More importantly, self-defeat arguments are typically quite weak, and Huemer’s are no exception. They typically rely on the halfway fallacy. Roughly, the halfway fallacy occurs when one grants that one’s opposition endorses some view, but then one presumes that one’s critic must be committed to some more fundamental or alternative view that conflicts with the view they hold or causes other philosophical problems for their views, such as (in this case) self-defeat. A foolish moral skeptic may endorse skepticism, and be committed to views inconsistent with their support for skepticism. If so, then by all means, criticize them for the inconsistency. But a skeptic can always go all in, rejecting whatever view they purportedly are committed to that results in a self-defeating position. Huemer or others may nevertheless reject skepticism as absurd and mistaken, but if so, it won’t be on terms the skeptic is obliged to accept. If a skeptic endorses both skepticism and some view that entails non-skepticism, their only recourse isn’t to abandon skepticism…they can simply reject the view that entails non-skepticism. As always, one philosopher’s modus ponens is another’s modus tollens. Nobody is obliged to accept self-defeat. There is always a way out.

Huemer doesn’t lay out a precise argument, but notes that skeptics appeal to two main arguments: that our moral intuitions are the result of unreliable processes (such as evolution or cultural traditions), and that there is so much moral disagreement that we should conclude people don’t have reliable moral intuitions. Both objections center on arguing that we don’t have reliable moral intuitions. So, we get something roughly like this:

P1. If we have unreliable moral intuitions, then we don’t have moral knowledge.

P2. We have unreliable moral intuitions.

C: We don’t have moral knowledge.

Huemer suggests that the very intuitions that would cause one to endorse P2 are themselves the result of unreliable intuitions, and that therefore if we were to suppose this argument was successful, we’d have to suppose that P2 was similarly threatened by an analogous argument directed against it. Perhaps something like this:

P1: If we have unreliable skeptical intuitions, the skepticism isn’t justified.

P2: We have unreliable skeptical intuitions.

C: Skepticism isn’t justified.

Consequently, the skeptic isn’t in a position to reject moral knowledge without undermining the very means that would justify doing so, so skepticism cannot get off the ground.

Skeptics have a few ways around this. First, they could show that they do have reliable intuitions, but moral realists don’t. This might involve arguing that there are unique problems with the factors that prompt realists to endorse moral realism.

Second, they could simply offer some other argument for denying that we have moral knowledge or that there are moral facts that don’t appeal to the distinctive unreliability of moral intuitions.

While I am sympathetic to both approaches, my own response is hard to categorize, and arguably is a mix of the two. I will develop this response in subsequent blog posts, but for now let’s take a tour of what Huemer has to say.

2.0 Unclear definition

According to Huemer:

Moral skeptics argue that either there aren’t any moral facts, or there are but we don’t know them. Why? Because our mechanisms for forming moral beliefs are not reliably truth-directed.

There are three problems with this definition. First, it’s not clear whether Huemer takes moral skeptics to specifically deny that there are stance-independent moral facts, or to deny that there are any moral facts at all, including stance-independent ones. In other words, does his account treat moral constructivists and relativists as skeptics, or not? It would be helpful, when discussing metaethics, to explicitly specify whether the moral facts you’re talking about are stance-independent or not. Even if you intend to include all moral facts, including stance-dependent ones, it’s still a good idea to explicitly specify this.

Second, he says “but we don’t know them.” This doesn’t specify whether the issue is that we currently don’t know what they are but could in principle acquire knowledge of them, or whether (for whatever reason) we are unable to acquire knowledge of them, either in principle or in practice. It might be helpful to clarify this, as well.

Third, is this: “Because our mechanisms for forming moral beliefs are not reliably truth-directed” …part of the definition of a moral skeptic? Or is Huemer defining a moral skeptic as anyone who denies there are moral facts or believes we don’t have knowledge of the facts, and then presenting a common characterization of why someone might be a moral skeptic? It’s not clear from his remarks.

Altogether, then, this is not a good way to start one’s response to moral skeptics. It’s not clear from Huemer’s remarks who counts as a moral skeptic. I think there are moral facts, but I don’t think there are stance-independent moral facts. I think we can have knowledge of these stance-dependent moral facts, but I almost certainly hold a different view about what “knowledge” is than Huemer. So am I a moral skeptic? I can’t tell from his description. His subsequent remarks sure do seem to indicate that I’m a moral skeptic, but as per his definition, it’s simply not clear.

So my first objection to Huemer’s post is that his remarks lack clarity. A clear criticism of moral skeptics should carve up the conceptual space using precise terminology that would allow readers to easily determine whether their positions fall in or outside the scope of his critique. While one might be able to, with some confidence, infer that their position does or doesn’t qualify as moral skepticism, this shouldn’t be necessary. His remarks should be clear enough that we can tell without having to do any guesswork.

I’m not the only person who found these remarks unclear. I ran a short survey on my YouTube page to see how others thought about these remarks:

Is this a representative sample? No, of course not. Voters are presumably people who subscribe to my channel, so they are likely going to be biased towards my perspective on these matters (though I didn’t give them any context that would further bias them). Nevertheless, I take it my audience can think for themselves and may very well have disagreed with me about this. Only many of them didn’t. While a substantial portion thought the arguments for skepticism were clearly not part of the definition (38%), there’s a good spread of other responses, with a majority thinking the meaning is not clear and a substantial subset thinking the arguments were part of the definition. I get that this is a blog post, and one could undoubtedly find ambiguities and lack of clarity in my blog posts, but that doesn’t mean it’s not worthwhile to point it out when it occurs.

3.0 Two arguments for moral skepticism

Huemer attributes two “sub-arguments” to moral skeptics:

a) Our moral intuitions are produced by something that is insensitive to moral truth, like natural selection, or the cultural traditions we happened to be born under.

b) There is so much disagreement among moral judgments that we have to conclude that humans can’t reliably judge morality.

I might endorse some version of (a), but I don’t endorse (b). Is that enough to make me a skeptic if Huemer’s conception of a moral skeptic requires one to endorse one or both arguments? I’m not sure. Even if it doesn’t, I am a moral antirealist, so my views are probably close enough. Either way, what Huemer suggests is that these considerations can be turned back on the skeptic:

I suggest that there are similar arguments debunking skepticism itself: skeptical beliefs are produced by unreliable processes that produce lots of disagreement.

This is why I provided the syllogisms above. Huemer could construct two arguments analogous to these but directed against moral skeptics. They’d probably take something like this form:

Argument from Unreliable Skeptical Intuitions

P1: If moral skepticism is produced by unreliable processes, then moral skepticism is unjustified.

P2: Moral skepticism is produced by unreliable processes.

C: Moral skepticism is unjustified.

Argument from Disagreement against Moral Skepticism

P1: If there is a lot of disagreement about moral skepticism, then moral skepticism is not justified.

P2: There is a lot of disagreement about moral skepticism.

C: Moral skepticism is not justified.

Arguments like these might work for a very flat-footed and crude type of moral skepticism. But they are more of a sketch than a detailed objection. It would be nice if we could neatly classify processes as “reliable” or “unreliable,” appeal only to those which are reliable, and call it a day. But human psychology isn’t as simple as that. Many of the judgments we make are subject to biasing factors, factors that potentially threaten their accuracy. But there are non-categorical questions: how much do these biasing factors threaten judgments that rely on a particular process? And under what circumstances do those threats result in false beliefs? Merely gesturing at some process, or judgment, and observing that biasing factors are involved isn’t enough to show that the process, or the judgments that result from its application to a particular case, have resulted in error. For comparison, consider memory. Is memory “reliable”? Well, what does that mean? The processes that are involved in memory mostly provide accurate information much of the time. How much of the time? Well, that depends on the person, the memory, and a host of distinctive factors. Nevertheless, memory is fallible and is often faulty. Its reliability is a matter of degree. There is no magic threshold for how often or to what extent it is accurate that would allow us to categorically declare it reliable or unreliable. Just so with any other psychological process.

Biasing factors are involved in virtually every judgment about everything. Simply noting that a person may have biases isn’t a good reason to conclude that whatever conclusions they’ve arrived at are the result of unreliable processes and that therefore they are mistaken. What we need is a broad, abductive case that appeals to all relevant considerations. Should we have more confidence, on reflection, in the case made by the skeptic or the case made by the realist? Answers to questions like this are not best resolved by attempting to force the messy weighing of competing theories through the strictures of deductive syllogisms, and simply saying “I’m biased? Well so are you!” isn’t going to necessarily end in the two sets of biases canceling each other out. After all, climate models are not perfectly reliable, but neither is reading tea leaves. That doesn’t mean that if climatologists appeal to previously-accurate climate models and oracles read the future in entrails that if their claims conflict then they epistemically cancel one another out.

4.0 General skepticism

According to Huemer,

Philosophers as a group have an extreme skeptical leaning, compared to people from other disciplines. For virtually anything that philosophers talk about (not just morality), one of the leading philosophical theories will be an extreme form of skepticism. This is not true of any other field of study that I know of.

I agree. This is true, as far as it goes. The problem is that Huemer is already moving towards establishing assumptions I don’t grant. This centers on his use of the term “extreme form of skepticism.” The term “extreme” is a highly loaded, normatively laden term. Whether a particular position is “extreme” or not is itself a potential matter of dispute. I simply don’t grant that there’s anything “extreme” about being a moral skeptic. On the contrary, I think moral realism is a completely absurd position for which there isn’t a single good argument. I find moral realism to be extreme, for much the same reason I’d think it extreme to believe in astrology or crystal magic.

Realists, and even many antirealists, are likely to disagree. That’s fine. I am not claiming there is a single standard for whether a position is extreme or not that we all should or must accept at the outset of inquiry. Whether a skeptical position is “extreme” depends on one’s other views (and of course on what one means by “extreme”). And since we don’t all share the same views about how well-supported a particular view is, we are not obliged to agree that any particular skeptical position is “extreme” or not. Yet Huemer compares skepticism about morality to skepticism about the existence of chemicals or rocks:

E.g., among chemists, a leading theory is not that there are no chemicals or that we know nothing about them. Among geologists, there is no theory that anyone takes seriously that says there are no rocks. Among art historians, a major view isn’t that there is no art or that it has no history.

Denial in these domains of inquiry of the phenomenon in question would involve denial of the very subject matter of the domain. Rocks, chemicals, and works of art just are the topics of study in these fields. However, moral philosophy isn’t the study of stance-independent moral facts. It’s the study of morality. Moral skeptics, and more generally, antirealists, do not necessarily deny that morality exists. Some might. But that’s not the only way to be a skeptic (or an antirealist). I don’t deny morality exists. I deny it involves stance-independence. A closer analogy to Huemer’s comparison to chemists and geologists would be a philosopher denying that there are such things as moral judgments, attitudes, positions, beliefs, values, language, and so on; that is, that as a social and psychological phenomenon, that morality literally doesn’t exist. Imagine denying that people say things like “That’s immoral.” Imagine denying that there is any such thing as a moral judgment, or a moral claim, or a moral value: that nobody, anywhere, has ever held a moral position on anything. That would be a more apt comparison to denying the existence of chemicals or rocks. But antirealists don’t (typically) do this.

Moral realism is not a required, operative assumption to do moral philosophy. It is a stance one takes within moral philosophy. Rejecting stance-independent moral facts would be closer to denying that life is characterized by Élan vital, a sui generis vital force that is present in all living things. Believing in vital force is not necessary to do biology. Likewise, believing in the quasi-mystical forces of goodness and badness that non-naturalist realists endorse simply isn’t necessary to do moral philosophy.

Even if Huemer doesn’t intend to suggest that we have equally good reason to endorse moral realism as we do the existence of chemicals or rocks, a reader could still plausibly interpret his remarks in this way. Of course, one can always intend a comparison to illustrate similarities in some set of features without implying similarities in others. I don’t want to be guilty of the overcomparison fallacy, myself. This occurs when you misinterpret the intent of a comparison. Suppose I want to compare lemons and bananas by noting that both are yellow. It would be foolish to say:

You can’t compare lemons and bananas! They’re completely different!

The fact that two things are different in various ways is completely irrelevant when it comes to assessing whether they are similar in some specified respect. Lemons and bananas both have the property yellow {Y}. It doesn’t matter if every other property in the respect set of properties for each were different:

Banana: {Y, A, B, C}

Lemon: {Y, D, E, F}

…It’d still be true to say “both have property {Y}”. If all Huemer has in mind by the comparison is that it’s as implausible to reject moral realism as it is to reject rocks and chemicals, then that’d be fine. But if that’s the goal, these are not very good choices. Both reflect the literal subject matter of the fields in question, while stance-independent moral facts simply are not the subject matter of moral philosophy. Stance-independent moral facts are a contentious, proposed set of phenomena that reflect one stance among philosophers. They aren’t the subject matter in the same way the Copenhagen interpretation of quantum mechanics isn’t the literal subject matter of physics. A physicist who rejects the Copenhagen interpretation isn’t denying the subject matter of physics, they’re just denying a specific interpretation of the nature of the subject matter.

Thus, even if the goal wasn’t to compare moral realism to geology and chemistry in respects other than it being similarly implausible to deny the phenomena in question, i.e., stance-independent moral facts, rocks, and chemistry, respectively, the fact that Huemer opted for phenomena that are the subject matter of the fields in question gives the impression that the same is true of moral philosophy, when this is not actually the case. This is thus at best unintentionally misleading and likely to exploit a greater implied similarity in the degree of absurdity of the skepticism.

Huemer’s comparison to the sciences is not ideal. Even if he only intended to offer analogous cases in which it’d be absurd to be skeptical, he chose these specific examples when he didn’t have to. If comparison with respect to characteristics the examples he gives share in common weren’t intended, it’d be very odd to choose such examples. In short, if the comparison were intended, it’s a mistake. If it wasn’t intended, then it’s misleading. Both are bad.